Mechanical ventilation (MV) is a cornerstone intervention in modern intensive and emergency care that is used to support hundreds of thousands of individuals each year while they cannot breathe independently. Unfortunately, this lifesaving intervention can cause harm: injudicious ventilator settings can promote lung injury, compromise circulatory stability, produce patient distress, stimulate an inflammatory response, and prolong the period of support required. Such adverse consequences of mechanical ventilation add to the burden of patient suffering, increase healthcare resource utilization, and compromise outcomes. Each year, relatively few practitioners receive detailed training in mechanical ventilation. Consequently, individuals requiring MV may be managed by practitioners with limited or spotty training in this intervention. In addition, the teaching of mechanical ventilation remains primarily a bedside exercise more akin to an apprenticeship than a systematic approach to mastery. New learners cannot practice extensively on actual patients, for ethical and practical reasons, and physiologically realistic alternatives are expensive and suffer from limited access. Moreover, exposure of the practitioner to the full spectrum of possible mechanical or physiologic derangements cannot be guaranteed. Those who are taught may not be taught well. Contemporary approaches to the assessment of expertise in mechanical ventilation are ill-suited for defining clinician practice patterns or competence in the context of a potentially harmful intervention for which any patient problem may have many possible solutions, the prevailing physiology is highly dynamic, and the clinician is reasoning in the setting of uncertainty. Similarly, ascertaining the extent to which a practitioner maintains her or his skills is hampered by the lack of clear metrics. We will address these three concerns by: 1. Developing an Internet deployable simulator in which learners confront and manage a carefully constructed population of virtual patients suffering from common cardiopulmonary maladies; 2. Developing and applying state of the art pattern recognition approaches to define practice patterns adopted by experts and novices as they manage virtual patients who require cardiopulmonary support; 3. Comparing the effectiveness of 3 different approaches to training: presentation based, simulation based, and simulation based training that adapts to focus on the learner's weak areas; 4. Elucidating the consequences of time away from training and simultaneous management of multiple virtual patients on provider performance. The proposed body of work will produce a cardiopulmonary management training tool deployable as freeware via the Internet, a database of expert and novice practice patterns, information regarding optimal training strategies, and data regarding compromise of provider performance by two relevant issues.